A DESIGN STUDY TO ENHANCE PERFORMANCE DASHBOARDS TO IMPROVE THE DECISION-MAKING PROCESS
Beem, Eric (2020-04-07)
A DESIGN STUDY TO ENHANCE PERFORMANCE DASHBOARDS TO IMPROVE THE DECISION-MAKING PROCESS
Beem, Eric
(07.04.2020)
Julkaisu on tekijänoikeussäännösten alainen. Teosta voi lukea ja tulostaa henkilökohtaista käyttöä varten. Käyttö kaupallisiin tarkoituksiin on kielletty.
avoin
Julkaisun pysyvä osoite on:
https://urn.fi/URN:NBN:fi-fe2020041719132
https://urn.fi/URN:NBN:fi-fe2020041719132
Tiivistelmä
Performance dashboards are tools that can be used to improve the decision making in an organisation (Henke et al., 2016). Nevertheless, organisations have trouble finding the right person to integrate and analyse the data in an organisation (Henke et al., 2016). This is not solely because the data analyst does not have the capabilities, but also because there is an information imbalance between the management board and the data engineer. Nowadays we live in a digital era and data plays an important role for organisations (McGee, Prusak and Pyburn, 1993). This thesis aims to solve this problem by creating an artefact to improve performance dashboards with explanatory business diagnoses. This will solve the imbalance between the management board and the data engineers and will improve the decisions in the organisation.
The first chapter is starts with the practical and scientific relevance and gives reasons why an artefact is needed. The research and sub questions are formulated, and the scope of this thesis is described. The second chapter focuses on the history of business intelligence (BI) and the role of BI in performance dashboard. Business intelligence and performance dashboards are related. Furthermore, the characteristics of performance dashboards and different performance dashboards are discussed. Multiple articles are combined to form four important characteristics for performance dashboards.
1. Flexibility: a performance dashboard needs to be easy to modify, used by multiple users and the ability to personalize the overview page.
2. Interactive: a performance dashboard needs to have the ability to drill down, monitor KPI’s and show not solely graphs.
3. Visual: a performance dashboard needs to give a visual overview of accurate data from the past and the present day in time.
4. External benchmarking: a performance dashboard needs to have the ability to compare the results with competitors and make prescriptive and predictive analysis based on the data.
This chapter ends with a comparison of different performance dashboards to find the most suitable tool for this research. Power BI is the most suitable tool for this research because it is easy to use and free. The focus of the third chapter is on the decision-making process. The articles of Mintzberg (1970) Endsley and Garland (2000) and Eppler and Mengis (2004) form the basis of this chapter. Information influences the decision-making process, but information can also lead to information overload (Eppler and Mengis, 2004). This chapter gives an overview of important factors in the decision-making process. These factors are used to improve the performance dashboard. Chapter four is about business diagnoses and explains the model of the artefact. The artefact is based on an article of Daniels and Feelders (2001). This article states that a good business diagnosis is based on six different steps.
1. determine the actual data (normalised/absolute and scaled or not scaled);
2. determine the reference data (normalised/absolute and scaled or not scaled);
3. get model relations from star scheme;
4. compute influence of reference data to determine causes;
5. filter causes to avoid information overload;
6. visual explanation tree of causes.
These steps are used to create the artefact. Chapter five analyses a new business diagnosis tool from Power BI. This tool is called the decomposition tree. This chapter finds out if it is useful for automated business diagnosis. The artefact is described in chapter six and different graphs and outcomes are displayed. The research ends with a conclusion about the advantages of the artefact, limitations and future research.
The first chapter is starts with the practical and scientific relevance and gives reasons why an artefact is needed. The research and sub questions are formulated, and the scope of this thesis is described. The second chapter focuses on the history of business intelligence (BI) and the role of BI in performance dashboard. Business intelligence and performance dashboards are related. Furthermore, the characteristics of performance dashboards and different performance dashboards are discussed. Multiple articles are combined to form four important characteristics for performance dashboards.
1. Flexibility: a performance dashboard needs to be easy to modify, used by multiple users and the ability to personalize the overview page.
2. Interactive: a performance dashboard needs to have the ability to drill down, monitor KPI’s and show not solely graphs.
3. Visual: a performance dashboard needs to give a visual overview of accurate data from the past and the present day in time.
4. External benchmarking: a performance dashboard needs to have the ability to compare the results with competitors and make prescriptive and predictive analysis based on the data.
This chapter ends with a comparison of different performance dashboards to find the most suitable tool for this research. Power BI is the most suitable tool for this research because it is easy to use and free. The focus of the third chapter is on the decision-making process. The articles of Mintzberg (1970) Endsley and Garland (2000) and Eppler and Mengis (2004) form the basis of this chapter. Information influences the decision-making process, but information can also lead to information overload (Eppler and Mengis, 2004). This chapter gives an overview of important factors in the decision-making process. These factors are used to improve the performance dashboard. Chapter four is about business diagnoses and explains the model of the artefact. The artefact is based on an article of Daniels and Feelders (2001). This article states that a good business diagnosis is based on six different steps.
1. determine the actual data (normalised/absolute and scaled or not scaled);
2. determine the reference data (normalised/absolute and scaled or not scaled);
3. get model relations from star scheme;
4. compute influence of reference data to determine causes;
5. filter causes to avoid information overload;
6. visual explanation tree of causes.
These steps are used to create the artefact. Chapter five analyses a new business diagnosis tool from Power BI. This tool is called the decomposition tree. This chapter finds out if it is useful for automated business diagnosis. The artefact is described in chapter six and different graphs and outcomes are displayed. The research ends with a conclusion about the advantages of the artefact, limitations and future research.